Conditional cash transfers: big in Latin America, small in Southeast Asia

The global debate on conditional cash transfers (CCTs), their design and impact on poverty alleviation is raging. More academic research is becoming available that shapes our knowledge about CCTs, their blessings and curses. The folks at the World Bank Development Impact Blog have posted a series of articles on the question whether conditions are a good thing?

 

While many Latin American and a number of African governments have successfully implemented large scale CCT programs, Southeast Asia has been a bit of a late bloomer. Pro-poor transfer programmes traditionally were (or still are) in-kind provisions, targeted subsidies or unconditional cash transfers.

The Indonesia government, for example, caused public outrage in 2005 when it cut back its generous fuel subsidies which were intended to benefit the poor, but, in effect, benefited those who were already rich enough to own private vehicles. Some of these resources were reallocated into a program called “Bantuan Langsung Tunai” (Direct cash assistance) which represented an unconditional household subsidy for poor families. The cash transfer was intended to counteract a potential increase in poverty incidence caused by the raise of gasoline prices.

In 2008, the Philippines launched a CCT programme called Pantawid Pamilyang Pilipino  (Building Bridges for the Filipino Family) which includes a 32 $ monthly cash grant to poor families on the condition that young children and pregnant mothers avail of preventive health care and that school-aged children attend school regularly. The programme reaches about 2.3 million Filipino households.

While impact studies show that conditions work, other evidence calls into question whether CCTs are superior to unconditional cash transfers (UCTs) under some circumstances. The World Bank Development Impact Blog provides insightful examples. They conclude:

 Imposing a condition distorts behavior. That may be a good thing, if it helps internalize market or information failures, or principal-agent problems within the household. However, both the existence of those problems and the fact that CCTs are the optimal policy response remain to be convincingly established. [...]

We know that many families do not take up a CCT offer, i.e. children do not attend school even in the presence of a CCT offer. This may be because the target population is heterogeneous in terms of ability, distance to school, or other determinants of the opportunity cost of schooling.’

The situation becomes more complicated when we consider the legislative processes needed to pass CCT programs. Elected representatives might not always be in favour of a new “incentivization’ if they will not get acknowledgement for the benefits. Political resistance to changing service provision for the poor in the Philippines illustrates this well.

• Some legislators will lose out during the budget allocation due to redirecting money from other less effective programs
• The process of distributing CCTs will be better targeted and thus legislators will have no discretion in who the beneficiaries will be
• A CCT program does not involve in-kind subsidy passed through the congressional district office such as rice, noodles, medicines or scholarships for which the poor had to queue up for hours in districit offices. The credit will then benefit the national leader (President Aquino III), instead the respective legislator.

Overall, CCTs can reduce dependency on local officials and reduce nepotism and corruption, rightly administered.

For some households, unconditional cash might be the better solution. On this, the Development Impact Blog concludes that CCTs are more effective in obtaining desired behaviour change, but they can undermine social protection for vulnerable households:

CCT programs create incentives for individuals to change their behaviors by denying transfers to those who don’t satisfy the conditions. However, at least some of these individuals come from vulnerable households and are also in need of income support. UCTs to such households can improve important outcomes even though they are not as successful in improving schooling outcomes as CCTs. Hence, while CCT programs may be more effective than UCTs in obtaining the desired behavior change, they can also undermine the social protection dimension of cash transfer programs.

Given the existing evidence and the different designs that are already out there, we are wondering what it is, that is holding back governments in Southeast Asia to experiment more with CCTs on a large scale? Yes, some of the side effects are unintentional, but overall the outcomes are beneficial. I will leave it at this with one last interesting finding: the puzzle of low elasticity to the transfer amount:

• comparative treatments in Cambodia, Morocco, Malawi and Nigeria have shown that even very low amounts yielded large improvements in school enrollement. The cash transfer consisted of 3-5% of household welfare.
• Isn’t that good news for policy makers?

For more specific information on ongoing impact evalutations of CCT programs around the world please refer to these additional sources:

World Bank Development Blog post on Conditional Cash Transfers (Part I)
Conditional Cash Transfer 2.0
World Bank Workshop Material: CCTs Second Generation of Evaluation